Conventional cordless telephone configurations include a handset coupled via radio connection with a base station. The base station is usually connected by wire to a traditional Public Switched Telephone Network (PSTN) or an Integrated Services Digital Network (ISDN). The development of new cordless standards which are based upon digital technology provides a broad spectrum of applications. Exemplary cordless applications include wireless Private Automatic Branch Exchange (PABX), wireless Local Area Network (LAN), Telepoint, and Radio Local Loop. Cordless standards include for example Digital Enhanced Cordless Telecommunications (DECT), Bluetooth, GSM, PHS, AMPS, IS54 or IS95. The digital cordless telephones represent a valid alternative to cellular phones in densely populated areas.
DECT is a cordless standard defined as a Multicarrier (MC), Time Division Multiple Access (TDMA)/Time Duplex Division (TDD) system. Time is divided in the DECT standard into frames of 10 ms. Each frame is divided into 24 full slots. The standard also allows for half slots and double slots of data.
In order to be able to support multiple channels, a DECT base station compresses and transmits 10 ms of speech during one full slot. This means that 10 ms of speech are actually sent over the radio in 416 μs. Every active connection makes use of two slots, one for receiving and one for transmitting. For example, if the slots in a DECT frame are numbered from 0 to 23, the first 12 slots (0-11) are used for transmission from the base station to the handset and the remaining slots are used for handset to base station transmission. A base station transmitting to a given handset in slot N receives from this handset in slot N plus 12, or in other words, half a frame later. Accordingly, a DECT base station is able to support up to 12 active voice connections at the same time.
The total number of bits within a conventional DECT slot is 480. With 24 slots and a 10 ms frame, a gross bit rate of 1.152 Mbits/s is provided. Once the DECT slot has been formatted, it is transmitted using one of 10 radio frequencies specified within the DECT standard. For example, the frequency band assigned to DECT in Europe is between 1,880 and 1,900 MHz, with a spacing of 1.728 kHz between adjacent frequencies. The transmission frequency for each channel is chosen dynamically based upon a Radio Signal Strength Indication (RSSI). Each active slot in the DECT frame may be transmitted and received on any of the 10 frequencies.
In exemplary communication systems, a first communication unit transmits voice samples to its counterpart unit. The TDMA structure is utilized to transmit and receive packets via a radio frequency (RF) channel to implement the exchange of voice samples between the two units. In typical digital communication systems, voice data is typically provided in an 8 kHz sample stream.
In a conventional DECT system, voice is typically carried over a fixed length packet using a single packet length cyclical speech buffer. Such service in DECT is called In_minimum_delay. Such guarantees that every slot in the TDMA structure carries the latest voice data. For this service, offsets within the speech buffer are used to address the latest speech samples to be used by the packet in the TDMA slot. If the voice is transmitted over two packets (e.g. in a handover case) both packets using different slots can use the same speech buffer but with different offsets. In DECT communications, there is no need for an additional buffer when switching form one slot (packet) to another slot (packet).
Referring to FIG. 1, an example of In_minimum_delay handover communications in a DECT system is illustrated. For example, during a handover between two slots (e.g. slot 0 and slot 4), voice data is utilized from a common buffer. However, the data communicated in each slot is different. For example, at the fixed part (FP) ADPCM data samples 0 to 39 are communicated to the portable part (PP) or handset in slot 0 (wherein offset is equal to zero). At the portable part, this data is forwarded to the ADPCM components starting with data 0 (wherein offset is equal to zero). At the time slot 4, the latest data arrives and is sent from the fixed part to the portable part in slot 4. Slot 4 includes data samples 6 to 39 (offset equals 6) which have also been sent in the previous slot 0 and the new ADPCM data 0-5 which arrived during the time between slot 0 and 4. At the portable part, slot 4 (offset equals six) will overwrite the data 6-39 received from slot 0. However, since this is the same ADPCM data there is no difference at the portable part ADPCM side for continuity. At the portable part, ADPCM data from slot 0 and slot 4 will be used depending on which slot has been received last. DECT In_minimum_delay allows voice data to be communicated between a fixed part and portable part irrespective of slot number.
The Bluetooth communication protocol standardizes data synchronization between disparate devices. The aim of Bluetooth communication protocol is to provide a single digital wireless protocol to address end-user problems arising from the proliferation of various mobile devices such as Smart Phones, Smart Pagers, hand held PCs, and Notebooks where it is desired to keep data consistent from one device to another.
According to the Bluetooth communication protocol or standard, it is described that voice packets, with different numbers of speech samples, can be used to transmit voice information. Depending on the number of speech samples in a packet, packets are spaced differently. Such permits selection between voice quality and amount of channel bandwidth used. A voice packet can be switched from one type to another type to permit dynamic selection between quality and bandwidth.
For example, if there are no desired bandwidth limitations, a high quality voice packet (HV1) with a minimum number of speech samples and a high number of added forward error correction (FEC) bits is used. However, when additional bandwidth is needed for other purposes (e.g. a second communication channel), a high quality voice packet is switched to a packet with a lower quality (HV2) and lower bandwidth requirements communicating a higher number of speech samples but a lower number of FEC bits.
Referring to FIG. 2, conventional voice packet switching according to the Bluetooth communication protocol is illustrated. Graph 2 corresponds to accessing data with respect to a high portion and a low portion of a first transmit buffer shown in FIG. 3. Data is written to the buffer from a data side and accessed from the buffer from a packet side in graph 2.
Graph 3 corresponds to accessing data with respect to a high portion and a low portion of a first receive buffer shown in FIG. 3. Data is read from the buffer from a data side and written to the buffer from a packet side 11 in graph 3.
Graph 4 corresponds to accessing data with respect to a high portion and a low portion of a second transmit buffer shown in FIG. 3. Data is written to the buffer from the port side and read from the buffer to a packet side in graph 4.
Graph 5 corresponds to accessing data with respect to a high portion and a low portion of a second receive buffer shown in FIG. 3. Data is written to the buffer from the packet side and read from the buffer to a port side in graph 5.
Graph 6 represents a TDMA frame structure comprising a plurality of TDD frames for the conventional operations.
Graph 7 represents transmit packets and graph 8 represents receive packets within a Bluetooth device. HV1 packets represent a minimum number of speech samples and a spacing of two slots for TDD frames −2, −1, 0.
For TDD frames 1-4, HV2 packets including a higher number of speech samples and a spacing of four slots are communicated. Such permits available bandwidth for other communications during TDD frames 2 and 4 as shown.
According to the Bluetooth communication protocol, three voice packets are defined using different repetition intervals. For packets HV1, TSCO (the repetition interval between packets) is equal to two and a new packet is started every two slots permitting one communication link. For HV2 packets, TSCO is equal to four wherein a new packet is started every fourth slot providing two communication links. For HV3 packets, TSCO is equal to six wherein a new packet is started every sixth slot providing three possible communication links.
Referring to FIG. 3, the exemplary Bluetooth buffer system is depicted. Two buffers configurations 9, 10 including respective buffer pairs 11, 12 are coupled with a packet composer 15 of an exemplary Bluetooth buffer system. Individual ones of buffers 11, 12 include respective buffer portions 13, 14. Buffer configuration 9 and buffer configuration 10 are coupled in parallel with packet composer 15 and synchronous I/O ports as shown. Conventional operations in FIG. 3 are described herein with reference to transmit operations.
Buffer configuration 9 corresponds to higher quality (HV1) communications while configuration 10 corresponds to lesser quality (HV2) communications. Buffer configuration 9 includes plural discrete buffers 11 individually having a size to hold a given number of data samples to be communicated in an HV1 packet. Buffer configuration 10 corresponds to lower quality communications and includes plural discrete buffers 12 individually having a size to hold more data samples than the number of data samples of buffers 11.
The buffers 11, 12 of respective configurations 9, 10 are used to communicate packets and switching between configurations 9, 10 is utilized to communicate packets having different numbers of data samples. During switching operations intermediate higher quality communications (HV1, and lower quality communications (HV2) buffers 11 and buffers 13 can not be properly aligned and accordingly errors result. Such interrupts speech data and is noticeable to a user.
Speech typically requires a synchronous continuous data channel. Any interruption in the speech data stream will be noticeable by the user in the form of clicks, missing speech, speech deformation, etc. Switching between different buffer configurations to communicate voice packets having different numbers of speech samples introduce interruptions in speech communications resulting in noticeable errors during the communications.